Chronic low-level lead poisoning is widespread, causing irreversible neurological problems in children. Assays of blood samples provide data on the amount of lead in circulation, but not on the accumulated lead concentration in the body - which is that contained in the skeleton. In vivo measurement of lead in the tibia by X-Ray fluorescence (XRF) is the present accepted method for determining lead in the skeleton. There are presently two XRF methods; one (the K XRF) uses the lead K X rays excited by the radioisotope source 109Cd while the other (the L XRF) uses the lead L X rays excited by the polarized X rays from an X-ray tube. Our previous research concentrated on improving the minimum detectable concentration (mdc) limits of both methods. This was accomplished by first developing a very comprehensive and accurate Monte Carlo model for simulating X-ray transport and the resulting spectra. It included L X rays, a detailed Compton scatter treatment, and polarized X rays. This model was then used to quantitatively determine errors in the two methods including variable skin covering thickness, variable and unknown sample position, and the presence of unknown lead concentration profiles in the bone. The Monte Carlo model was then used to investigate various methods of improving the minimum detectable lead limits. The primary improvement in both cases (a factor of about three in the reduction of statistical uncertainties) was found in using the entire spectral information rather than just the characteristic K and L peaks. The approach used is called the Monte Carlo - Library Least-Squares (MCLLS) approach. It is based on determining the elemental spectral libraries for any sample composition and analysis geometry by Monte Carlo simulation and then using these libraries in a linear least-squares analysis of the unknown sample spectrum for lead amount. An even more comprehensive and promising approach was identified and investigated in a very preliminary way. That approach is the combined use of the K and L methods by the MCLLS approach; thereby providing a direct basis for eliminating or at least minimizing the previously identified errors even more. The proposed method would use the radioisotope source 109Cd for excitation of both K and L X rays via the 88 keV gamma ray and silver K X rays, respectively. The major advantages of this approach are: (1) all the available spectral information is used in an optimum way, (2) one can determine if a lead concentration profile exists since the K and L X rays penetrate different bone thicknesses, and (3) the possibility exists that the spectral data taken can be used to eliminate the effect of unknown sample position variations and other patient-dependent variables that interfere with the lead measurement. The development and testing of this approach is the research proposed here.